In September 2024, users discovered that LinkedIn had quietly updated its privacy settings to automatically opt all members into allowing their personal data — including posts, articles, and activity — to be used to train its AI models. The setting was enabled by default. To opt out, users had to navigate several menus in a non-obvious location. Privacy advocates in the EU immediately flagged it as likely illegal under GDPR. The UK's Information Commissioner's Office launched a probe. LinkedIn said it had 'paused' training on UK data, but the damage was done: hundreds of millions of users had been silently enrolled without meaningful consent.